HRCT LUNG FINDINGS IN POST CHEMOTHERAPY PATIENTS ... · niques are required due to the limitations of the pattern approach on chest radiography. High-resolution computed tomography

JPMI VOL. 29 NO. 4 243

HRCT LUNG FINDINGS IN POST CHEMOTHERAPY PATIENTS - EMERGING CHALLENGE FOR RADIOLOGISTS

Ummara Siddique Umer1, Shahjehan Alam2, Syed Ghuas3, Seema Gul4, Qurat Ul Ain Arif5

ORIGINAL ARTICLE

INTRODUCTIONPulmonary diseases due to drug toxicity are com-

mon but probablynot detected and reported frequent-ly1. Thus, the true incidence of drug-induced respiratory disease is unknown, although it is estimated to be less than 10%2.

Chemo-therapeutic agents besides other drugs might be responsible for adversely affecting the lungs3-6.As their use is on the rise, physicians as well as radiolo-gists will be facing a huge challenge.Symptoms are usu-ally non-specific. Patients may present with shortness of breath, fever and cough which can appeareven after a significant period of time has been lapsed after taking the drug. Alowthreshold for its diagnosis should be kept asit can mimic infection andunderlying lung disease.

The most frequent radiologic features associated with chemotherapy are pulmonary infiltrates. However, in early stages, it can be asymptomatic. The ground-

glass infiltrates should not be ignored ormis-interpret-ed as tumor progression7. Chemotherapy associated pulmonary infiltrates may develop due to a variety of reasons. Infection, inflammation, cardiogenic and non-cardiogenic interstitial edema are considered to be significant. Knowledge regarding chemotherapy agents which are mainly responsiblefor pulmonary tox-icity may be helpful in assessing the most likely cause. Some of the older chemotherapy agents (bleomycin and carmustine) can lead to dose dependent pulmo-nary toxicity. Others (cyclophosphamide, busulfan, and carmustine) can cause it even years after completion of therapy8.

Standard plain chest radiograph is initially consid-ered for assessing drug associated pulmonary toxicity in most of the patients.However, other imaging tech-niques are required due to the limitations of the pattern approach on chest radiography.

High-resolution computed tomography (HRCT) can

This article may be cited as: Umer US, Alam S, Ghuas S, Gul S, Arif QA. HRCT lung findings inpost chemotherapy patients - Emerging challenge for radiologists. J Postgrad Med Inst 2015; 29(4): 243-51.

ABSTRACTObjective: To characterize, diagnose and to differentiate various HRCT mani-festations of lung abnormalities in post chemotherapy patients.

Methodology: This was a retrospective study of 50 patients conducted at Ra-diology department of Rehman Medical Institute, Peshawar. Duration of study was 6months i.e from April 2013 to September 2013. Patients were investigat-ed using 128-slice Multidetector Computed tomography (MDCT) scanner in the Radiology department of Rehman Medical Institute Peshawar.0.5mm re-constructed images in lung window and 3mm images in mediastinal window were viewed on workstation in axial, coronal and sagittal planes. The data was processed using Microsoft excel 2007.

Results: A total of 50 patients were included. Age of the patients ranged from 6 to 70 years with a mean age of 35 years. In our study, we found five radiologic patterns on CT scan; (1)non-specific ground-glass attenuation 17(34%),(2) patchy distribution of ground-glass attenuation accompanied by interlobular septal thickening 7(14%), (3)multifocal areas of airspace con-solidation 7(14%),(4)extensive bilateral ground-glass attenuation or airspace consolidations with traction bronchiectasis 4(8%), and (5) nodules of variable sizes randomly distributed in both lungs 15(30%).

Conclusion: The most common pattern was found to be patchy areas of ground-glass attenuation. Pulmonary diseases that are induced by chemo-therapy represent particular challenges for radiologists due to non-specific and atypical imaging features.

Key Words: Post chemotherapy lungs, High resolution computed tomogra-phy (HRCT), Multidetector computed tomography (MDCT)

1-5 Department of Radiology, Rehman Medical Institute, Peshawar - Pakistan.Address for correspondence:Dr. Ummara Siddique UmerSenior Registrar,Department of Radiology, Rehman Medical Institute, Peshawar - Pakistan.E-mail: [email protected] Received:November 04, 2014Date Revised:November 12, 2015Date Accepted:December 14, 2015

HRCT LUNG FINDINGS INPOST CHEMOTHERAPY PATIENTS - EMERGING CHALLENGE FOR RADIOLOGISTS


aid in early detection of lung toxicity, when it is still re-versible. It can also help in differentiating drug toxicity from other lung pathology9. In symptomatic patients with normal or less distinct findings on chest radiogra-phy, HRCT can playkeyrole in the delineation of paren-chymal abnormalities10,11.

METHODOLOGYThis was a retrospective study of 50 patients con-

ducted at Radiology department of Rehman Medical Institute, Peshawar. Duration of study was 6months i.e from April 2013 to September 2013. Approval from Reh-man medical research committee of ethics and research was taken. Enrollment of OPD and indoor post chemo-therapy patients coming for HRCT chest with age range of 6- 70 years (Mean age: 35 yrs) was done. Informed consent was undertaken from the patient. Nature and duration of the procedure, benefits and involved docu-mented risks of the intravenous contrast were narrated in an easy /native language to the patients. HRCT chest was performed on a 128 slice Toshiba MDCT scanner (CT techniques: 128 slicer CT scanner, Toshiba Aquil-ion, Tokyo, 120kvp / 300mA, 150mAs, 500msec, 0.5mm thickness). Additional complimentary prone scanning was done when indicated.All 50 patients which were in-

cluded in the study were proven cases of malignancy on the basis of Biopsy and on chemotherapeutic treatment cycles. All data was entered and analyzed using SPSS version 10.0 and processed using Microsoft excel 2007.

Retrospective review of HRCT scans on vitrea work-station was done by four qualified radiologists for pres-ence or absence of ground glass haze, airspace con-solidation, reticulation and honeycombing, extent and zonal distribution of HRCT findings, severity of bron-chiolectasis / bronchiectasis, architectural distortion or ancillary findings such as lymphadenopathy and associ-ated pleural or cardiac changes.

RESULTSA total of 50 patients were included. Age of the pa-

tients ranged from 6 to 70 years with a mean age of 35 years. In our study, we found five radiologic patterns on CT scan. Their relative frequencies are shown in (Table1.)

Findings suggesting classical Tuberculosis(TB) were seen in 5 patients, out of which 4 had normal back-ground whereas 1 patient had background lung chang-es of previous TB. Two patients had imaging features characteristic of Invasive Aspergillosis. (Table2.)

Table 1: HRCT patterns in post chemotherapy lungsS.no HRCT patterns Number of patients (%)1 Patchy ground glass haze(GGH) 17 (34%)2 Patchy GGH with interlobular septal thickening 7(14%)3 Multifocal areas of consolidation 7(14%)4 Extensive bilateral GGH or consolidations with traction bronchiectasis (

like in NSIP) 4(8%)

5 Nodules of variable sizes randomly distributed in both lungs 15(30%)

Table 2: Lung Disease patterns in post chemotherapy patientsDisease pattern Number of patientsTuberculosis 5 (10%)Invasive Aspergillosis 2 (4%)UIP 3 (6%)Lymphangitis 3 (6%)Radiation Induced Fibrosis 1 (2%)Metastatic Nodules 9 (18%)Granulomas 6 (12%)Nonspecific Patchy GGH only 17 (14%)NSIP/Chemotherapy induced Toxicity 4 (8%)



Figure 1: (a&b) HRCT image of a 30 year old male showing bilateral patchy areas of ground glass attenuation (arrows) with visualization of underlying normal vessels and

positive dark bronchus sign(fig1b) 1c: Drug induced toxicity in a 48 years old CA breast patient on chemotherapy. CT chest shows

patchy areas of nonspecific ground glass haze. Nodular mass noted in right breast.

Figure 2: CT chest sagittal image showing patchy areas of ground glass haze(GGH) more so in dependent segments of lungs, in keeping with pulmonary edema.

1a 1b

1c



Figure 3: (3a) HRCT image showing patchy areas of ground glass haze interspersed with septal thickening (crazy-paving pattern).

(3b) HRCT chest of another patient showing left upper lobe and lingular patchy ground glass haze and septal thickening. Few consolidating opacities also seen anteriorly.Tiny cen-trilobular bullae seen in both lungs, which are more clearly seen on left side due to back-

ground of GGH.

Figure 4: (4a)HRCT image of a 68 years old male showing both fine and nodular thicken-ings of interstitial tissue septa (arrows),in keeping with lymphangitis.(5b) HRCT image of

another patient shows a small patch of septal thickening (arrows) in right lung base, highly suspicious for lymphangitic spread of disease (lymphangitis).

3a 3b

4a 4b

DISCUSSION

Approximately 10% of patients receiving cytotox-ic drugs develop drug-induced lung reactions12,13. Any chemotherapeutic drug can adversely affect the lung, but the most commonly implicated drugs associated withlung toxicity are bleomycin, methotrexate, carmus-tine, busulfan, and cyclophosphamide14-18.

Currently,HRCT is considered the best non-invasive method to assess the presence of drug-induced lung disease. It may show abnormalities even in patients with normal radiographs19-21. An accuracy of 45% has been reported for HRCT by Cleverly et al22. Therefore, to pre-dict the histological patterns in drug-induced lung dis-eases,HRCT may have limited ability22. However,HRCT is of great value in excluding alternative diagnoses and in monitoring response to treatments.



Figure 5: HRCT chest images showing classic UIP pattern of bilateral sub-pleural areas of septal thickening, reticulations, clusters of tiny air cysts, honeycombing and tractional

bronchiectasis with apicobasal predominance.

Figure 6: (a&b) CT chest of a patient with Right lung carcinoma showing a consolidation in right upper lobe.Mediastinal windowreveals fluid filled bronchi in the consolidated lung.

Lung window shows peri-consolidation areas of septal thickening (arrows), which are suggesting regional spread of disease. Right sided aortic arch seen with aberrant

left subclavian artery.

5a

6a

5b

6b

(1) Patchy areas of ground glass haze (GGH)

GGH is the slight increase of pulmonary attenu-ation, which permits seeing the underlying vessels and walls of the bronchi23 (dark bronchus seen). (Fig-ure.1) Its cause could be edema, cells like neutro-phils or even early changes of fibrosis in which the underlying pulmonary alteration is below the limit of resolution of the HRCT image. Though non-spe-

cific in itself, the sign is always very significant23. In our study patchy ground glass haze areas were di-vided into two groups, Patients whose GGH shifted on prone scanning were considered as dependent GGH like in pulmonary edema which is a side effect of long standing chemotherapy or illness(Figure.2). Inpatients whose GGH didn’t shift with prone imaging were la-belled as nonspecific GGH. These were the cases that were advised follow up HRCT.



Figure 7a: CT chest of a 41 years old female,known case of CA stomach, showing small thick walled cavity and tree in bud branching opacities in posterior segment of right

upper lobe. These findings are classical for active endobronchial Tuberculosis.

Figure 7b: (1&2) HRCT chest axial and coronal images of another patient patient showing consolidations, multiple nodules and tree-in-bud like branching nodular

opacities(arrows) in active endobronchial tuberculosis.

7a

7b1 7b2

(2)Septal thickening with or without GGH.

The combination of ground-glass opacities and thickening of interstitial septa can produce the cra-zy-paving pattern(Figure.3). The thickening of inter-stitial tissue septa could be either fine or nodular. The nodular thickening associated with GGH needs special concern to rule out lymphangitis carcinoma-tosa (Figure.4) and further histopathological workup should be considered. Fine septal thickening can be seen with tumour progression but it is very rare and mostly could be due to pulmonary edema orearly changes of fibrosis with alveolitis .The CT findings in patients with leukemia consist mainly of ground-glass attenuation, centrilobular nodules and thickening of the bronchovascular bundles in the peripheral lung24.

In our study 02 patients showed HRCT findings of septal thickening with GGH, and for which strong concern for lymhphangitis was raised. Patients with septal thicken-ing without GGH were grouped according to the zonal predominance. Subpleural septal thickening with a pre-dominant apicobasal gradientwaslabeled as usual inter-stitial pneumonia (UIP)(Figure.5). In our study we only had two patients with classical UIP like pattern. These were considered as incidental. There was a patient with focal patch of septal thickening in right anterior lung, which was considered as radiation induced fibrosis.

(3)Patchy areas of consolidations

Consolidations are areas of increased air space attenuation which obscure the underlying vascular



markings23 (Figure.6). These could be with or without air bronchograms. In our study, patchy consolidations were seen in 14% patients.In post chemotherapy pa-tients, consolidating opacities in lungs could be infec-tive in etiology i.e. with superadded infection (Bacterial, Mycobacterial – TB, Fungal), might be Lymphomatous deposits in cases ofLymphoma, Hypersensitivity pneu-monitis due to drug induced hypersensitivity or of malignant etiology. In our series, consolidations were seen in lungs of 5 patients with associated branching nodular opacities and small cavities (Figure.7), which is a classical finding of pulmonary tuberculosis.In two cases,there were periconsolidation ground glass halos which strongly represented invasive aspergillosis. The other differential considered in these cases wascrypto-genic organizing pneumonia (COP).

(4) Extensive bilateral ground-glass attenuation or airspace consolidations with traction bronchiec-tasis (like in NSIP):

The chronic form of chemotherapy induced lung tox-icity pattern manifests as a pulmonary fibrotic process,

often with no accompanying eosinophilia. In our series, 4 patients had classical imaging appearance of nonspe-cific interstitial pneumonia/Chemotherapy induced tox-icity with patchy areas of extensive GGH, consolidations and septal thickening with tractional bronchiectasis with zonal apico-basal predominance (Figure 8).

(5) Multiple nodules:

In our study we characterized pulmonary nodules into three groups: i) Pulmonary granulomas. These are com-posed of macrophages reacting to various chemother-apeutic drugs, such as methotrexate and nitrofurantion. These can also be post infective or post tuberculous and not associated with chemotherapy effects. ii)Metastatic nodules: classic imaging features of metastatic nodules is a long list.Main features seen in our patients were positive feeding vessel sign, subpleural location, lobular spiculated margins and interval size and number prog-ress.iii)Nodules with Equivocal features. These were sug-gested follow up. These nodules should be correlated with PET scan. There were few limitations in our study. Limited number of the study enrolled patients came for

Figure 8: (a,b&c) HRCT chest images showing imaging appearance of nonspecific interstitial pneumonia/ drug induced toxicity. Bilateral patchy areas of ground

glass haze, sub pleural reticulations, septal thickening and tractional bronchiecta-sis seen with apicobasal predominance. Septal thickening with tractional bronchi-

ectasis suggest early fibrotic changes, which are mostly irreversible.

8a 8b

8c



follow up HRCT.No correlation of findings and changes on HRCT scan with clinical and functional parameters in most of the patients. No consideration of changes during the whole disease course that is, only compari-son of initial and the recent follow up exams.They char-acterized CT findings into four groups whereas we di-vided our HRCT findings into five groups. Nodules are an important CT feature for assessing post chemother-apy lungs and should be a part of check-list in lungs. In our study most common finding was found to be patchy ground glass haze which was the same as study conducted by Endo et al in Japan. According to Santi-ago et al, Ground-glass haze is nonspecific, but high-ly significant finding since 60-80% of patients have an active and potentially treatable lung disease. Previous studies have evaluated the prognostic significance of HRCT in chemotherapy induced lung disease patterns.According to Wells and Gay et al, Response to treat-ment is better in patients with GGH and is correlated with the extent of GGH on HRCT25. Patients with more extensive fibrosis on HRCT have a worse prognosis26. Usually, GGH indicates active disease that is potentially reversible with the appropriate treatment.

Akira et al reported diffuse or multi-focal ground-glass opacities with intralobular interstitial thickening as the mainfeature in chemotherapy associated pneumo-nitis. On the contrary, in antibiotic-induced pneumo-nitis, predominant radiographic findings were patchy ground-glass opacities with centrilobular opacities and interlobular septal lines27.

Although HRCT findings are usually non-specific, diagnosis can be arrived by identifying the most com-mon histo-pathologic and radiologic manifestations of drug-induced lung injury28.

Ground-glass opacities and mild fibrosis are fre-quently ignored on radiography29. Patients with non-specific patchy areas of ground glass haze on HRCT and that too without septal thickening (Figure.2c) must not be ignored as most of the times these early changes respond to treatment. Follow up scan must strongly be advised.

CONCLUSIONThe most common pattern was found to be patchy

areas of ground-glass attenuation. Pulmonary diseases that are induced by chemotherapy represent particular challenges for radiologists due to nonspecific and atyp-ical imaging features. Patients with nonspecific patchy areas of ground glass haze and that too without septal thickening must not be ignored as most of the times these early changes respond to treatment. Follow up scan must strongly be advised.

REFERENCES1. Cooper JA Jr. Drug-induced lung disease. Adv Intern Med

1997; 42:231-68.

2. Camus P, Bonniaud P, Fanton A, Camus C, Baudaun N, Foucher P. Drug-induced and iatrogenic infiltrative lung disease. Clin Chest Med 2004; 25:479-519

3. Meyers JL. Pathology of drug-induced lung disease. In: Katzenstein AA, Askin FB, eds. Katzenstein and Askin’s surgical pathology of non-neoplastic lung disease. 3rd Ed. Philadelphia: Saunders; 1997:81-111.

4. Cooper JA Jr, White DA, Matthay RA. Drug-induced pul-monary disease. Cytotoxic drugs. Am Rev Respir Dis 1986; 133:321-40.

5. Ginsberg SJ, Comis RL. The pulmonary toxicity of antineo-plastic agents. Semin Oncol 1982; 9:34-51.

6. Sostman H, Matthay RA, Putman CE, Smith GJ. Metho-trexate-induced pneumonitis. Medicine (Baltimore) 1976; 55:371-88.

7. Torrisi JM, Schwartz LH, Gollub MJ, Ginsberg MS, Bosl GJ, Hricak H. CT findings of chemotherapy-induced toxicity: what radiologists need to know about the clinical and ra-diologic manifestations of chemotherapy toxicity. Radiol-ogy 2011; 258:41-56.

8. Meadors M, Floyd J, Perry MC. Pulmonary toxicity of che-motherapy. Semin Oncol 2006; 33:98–105.

9. Graves MW, Kiratli PO, Mozley D, et al. Scintigraphic di-agnosis of a right to left shunt in end-stage lung disease. Respir Med 2003; 97:549-54.

10. Shah RM, Miller W Jr. Widespread ground-glass opacity of the lung in consecutive patients undergoing CT: Does lobular distribution assist diagnosis?. AJR Am J Roentge-nol 2003; 180:965-8.

11. Duran I, Siu LL, Oza AM, et al. Characterisation of the lung toxicity of the cell cycle inhibitor temsirolimus. Eur J Can-cer 2006; 42:1875–80

12. Endo M, Johkoh T, Kimura K, Yamamoto N. Imaging of ge-fitinib-related interstitial lung disease: multi-institutional analysis by the West Japan Thoracic Oncology Group. Lung Cancer 2006; 52:135–40.

13. Rosenow EC 3rd, Limper AH. Drug-induced pulmonary disease. Semin Respir Infect 1995; 10:86-95.

14. Cannon GW. Methotrexate pulmonary toxicity. Rheum Dis Clin North Am 1997; 23:917-37.

15. Alexandrescu DT, Dutcher JP, O’Boyle K, Albulak M, Oiseth S, Wiernik PH. Fatal intra-alveolar hemorrhage after rit-uximab in a patient with non-Hodgkin lymphoma. Leuk Lymphoma 2004; 45:2321-5.



16. Garrido M, O’Brien A, González S, Clavero JM, Orellana E. Cryptogenic organizing pneumonitis during oxaliplatin chemotherapy for colorectal cancer: case report. Chest 2007; 132:1997-9.

17. Ohnishi K, Sakai F, Kudoh S, Ohno R. Twenty-seven cases of drug-induced interstitial lung disease associated with imatinibmesylate. Leukemia 2006; 20:1162-4.

18. Buttin BM, Moore MJ. Thalidomide-induced reversible in-terstitial pneumonitis in a patient with recurrent ovarian cancer. Gynecol Oncol 2008; 111:546-8.

19. Silva CI, Müller NL. Drug-induced lung diseases: most common reaction patterns and corresponding high-reso-lution CT manifestations. Semin Ultrasound CT MR 2006; 27:111-6.

20. Myers JL, Limper AH, Swensen SJ. Drug-induced lung dis-ease: a pragmatic classification incorporating HRCT ap-pearances. Semin Respir Crit Care Med 2003; 24:445-54.

21. Ellis SJ, Cleverley JR, Müller NL. Drug-induced lung dis-ease: high-resolution CT findings. AJR Am J Roentgenol 2000; 175:1019-24.

22. Cleverley JR, Screaton NJ, Hiorns MP, Flint JD, Müller NL. Drug-induced lung disease: high-resolution CT and histo-logical findings. Clin Radiol 2002; 57:292-9.

23. Zompatori M, Rimondi MR. Diffuse ground-glass opacity of the lung. A guide to interpreting the high-resolution computed tomographic (HRCT) picture. Radiol Med 1994; 88:576-81.

24. Okada F, Ando Y, Kondo Y, Matsumoto S, Maeda T, Mori H. Thoracic CT findings of adult T-cell leukemia or lym-phoma. AJR Am J Roentgenol 2004; 182:761–7.

25. Wells AU, Hansell DM, Corrin B, Harrison NK, Goldstraw P, Black CM, et al. High resolution computed tomography as a predictor of lung histology in systemic sclerosis. Thorax 1992; 47:508-12.

26. Gay SE, Kazerooni EA, Toews GB, Lynch JP 3rd, Gross BH, Cascade PN, et al. Idiopathic pulmonary fibrosis: predict-ing response to therapy and survival. Am J Respir Crit Care Med 1998; 157:1063-72.

27. Akira M, Ishikawa H, Yamamoto S. Drug-induced pneu-monitis: thin-section CT findings in 60 patients. Radiology 2002; 224:852-60.

28. Erasmus JJ, McAdams HP, Rossi SE. High-resolution CT of drug-induced lung disease. Radiol Clin North Am 2002; 40:61-72.

29. Padley SP, Adler B, Hansell DM, Muller NL. High-resolu-tion computed tomography of drug-induced lung dis-ease. Clin Radiol 1992; 46:232-6.

CONTRIBUTORSUSU Conceived the idea, planned the study, and

drafted the manuscript. SA helped acquisition of data and did statistical analysis. SG and QUAA drafted and critically revised the manuscript. All authors contrib-uted significantly to the submitted manuscript.

Documents

HRCT LUNG FINDINGS IN POST CHEMOTHERAPY PATIENTS ... · niques are required due to the limitations of the pattern approach on chest radiography. High-resolution computed tomography